| Author |
Message
|
| sean_m_hogan |
 Posted: Thu Apr 05, 2012 1:55 am Post subject: MQ 6.0.2.8 : strmqm : Compcode=0 Reason=2059 |
 |
|
Newbie
Joined: 28 Sep 2006
Posts: 8
Location: UK
|
Hi
How to start up queue manager ?
Environment : MQ6 on AIX5.3
Was working on 6.0.2.11 but regressed on instruction as there are perceived performance issues (though I feel the problem with that might lie with the configuration).
Back to 6.0.0.9 then installed 6.0.2.8
Now when the HACMP is cluster is brought up the queue manager stays down. Experimenting with strmqm under user mqm produces the reason code 2059. No FDC files produced and nothing in the AMQERR*.LOG files.
Prior to posting this query I had googled and checked elsewhere on this excellent site.
Thanks in advance
Sean Hogan
_________________
Sean M. Hogan
Rusty MQ user - OK at ticking multiple-choice |
|
| Back to top |
|
 |
| exerk |
| Posted: Thu Apr 05, 2012 2:25 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
What's in the HACMP logs?
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Thu Apr 05, 2012 4:00 am Post subject: Re: MQ 6.0.2.8 : strmqm : Compcode=0 Reason=2059 |
|
|
Poobah
Joined: 15 May 2001
Posts: 7716
|
| sean_m_hogan wrote: |
Now when the HACMP is cluster is brought up the queue manager stays down. |
Are all the files really there and available? On the server where you are getting the 2059 from the strmqm command, can you navigate into that QM's directory and view the queue files, the log files?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| sean_m_hogan |
| Posted: Tue Apr 10, 2012 12:48 am Post subject: |
|
|
Newbie
Joined: 28 Sep 2006
Posts: 8
Location: UK
|
Apologies for tardy update. Lost my log-on password.
No FDC files generated. No update to AMQ*LOG files.
The files structure on the cluster appear
Maybe I should run 'installp -C' to clear up following a regression from 6.0.2.11 to 6.0.2.8 ?
# lslpp -al | grep 6.0.2.11 <== nothing returned
# lslpp -al | grep 6.0.2.8 <== Filesets returned
& 'dspmqver' shows 6.0.2.8
Might there be some element of 6.0.2.11 lurking which hinders MQ running as 6.0.2.8 ?
Should I be delving into the ODM ?
_________________
Sean M. Hogan
Rusty MQ user - OK at ticking multiple-choice |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Tue Apr 10, 2012 4:06 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7716
|
What happens if you bring the disk group(s) online that hold the Queue Manager's queues and logs, and then instead of using the clustering software to (attempt to) online the Queue Manager, you instead log in as mqm on the active server and issue the strmqm command?
This little test would help identify where the problem lies. If your strmqm command works, then the piece that onlines/offlines the QM in your cluster needs to be looked at. If the strmqm command fails or hangs, then no wonder the cluster can't online it if you can't. At that point focus on why the simple strmqm command fails / hangs. On the IBM MQ Support site, in the Must Gather sections, is a scenario of what info to gather if your commands hang or fail but no errors are otherwise generated.
What does the dspmq command show at various points during the above?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| sean_m_hogan |
| Posted: Tue Apr 10, 2012 4:48 am Post subject: |
|
|
Newbie
Joined: 28 Sep 2006
Posts: 8
Location: UK
|
The MQ binaries are stored locally so each physical server in the cluster is separately regressed/upgraded. The queue manager structure is stored on clustered disks.
So a simple crmqm TEST_MQ results in a queue manager with all qmgr objects local to that machine.
It's 'strmqm TEST_MQ' where I hit a snag.
Just awaiting formal permission to resume this..environment is preproduction...
_________________
Sean M. Hogan
Rusty MQ user - OK at ticking multiple-choice |
|
| Back to top |
|
|
| sean_m_hogan |
| Posted: Thu Apr 12, 2012 5:14 am Post subject: |
|
|
Newbie
Joined: 28 Sep 2006
Posts: 8
Location: UK
|
I rolled back the MQ versions using the AIX admin tool smitty, i.e. :
- rejected (applied) filesets for 6.0.2.11
- Same for 6.0.2.8
- Un-installed MQ
Then :
- Installed MQ6.0 base and omitted Java 1.4 Java 3D to overcome the failed dependencies on OpenGL.rte base & root 5.1.0.50
- Reapplied filesets update for 6.0.2.8
Can start a test queue manager.
Then restarted the HACMP cluster services, failed over MQ (& message broker) from the other node (running 6.0.2.11) and following a repeat of the above steps failed back.
So the exact cause of the problem remains elusive so I had to resort to this extreme resolution. At least we didn't lose the message broker queues and channels when MQ was un-installed.
_________________
Sean M. Hogan
Rusty MQ user - OK at ticking multiple-choice |
|
| Back to top |
|
|
|
|
